Abstract Body (Do not enter title and authors here): Background: The risk of cardiovascular disease (CVD) varies among patients with diabetes. Evidence is insufficient for comparing coronary artery calcium (CAC) and measures of renal dysfunction in reclassifying CVD risk among these patients.
Hypothesis: This study assesses whether CAC enhances discrimination and net reclassification improvement (NRI) compared to measures of renal dysfunction in persons with diabetes.
Methods: We used Cox proportional hazard models to compare the performance of CAC and measures of renal dysfunction in predicting CVD in the Multi-Ethnic Study of Atherosclerosis (MESA). Models adjusted for age, sex, race, ever smoking, body mass index, systolic blood pressure, total cholesterol, high-density cholesterol, triglycerides, A1c, fasting blood sugar, hypertension medication use, and insulin use. Measures of renal dysfunction included estimated glomerular filtration rate (eGFR) and urine albumin-creatinine ratio (UACR), assessed individually or as composite variables (eGFR <60 mL/min/1.73 m² or UACR >30 mg/g). Improvements in c-statistics and the 10-year area under the curve (AUC) were calculated.
Results: A total of 839 participants, with a mean age of 64.5 ± 9 years, 396 female (47.2%), 155 White (18.4%), 321 Black (38%), 260 Hispanic (31%), and 103 Chinese (12%), were included. Of the total, 528 (63%) had CAC >0, 139 (17%) had eGFR <60 mL/min/1.73 m², and 232 (28%) had UACR >30 mg/g. Patients were followed up for a mean of 13.4 years. CAC, eGFR, and UACR were associated with an increased risk of CVD in all predefined models. CVD risk increased incrementally with the increase in CAC category (Table). The addition of CAC significantly increased AUC in Model 1 with baseline variables (0.680 vs. 0.726, p=0.002), Model 2 with individual measures of renal dysfunction (0.702 vs. 0.745 p=0.003), and Model 3 with composite measures of renal dysfunction (0.706 vs. 0.746, p=0.004).
Conclusion: CAC reclassifies CVD risk beyond measures of renal dysfunction and may be used to identify higher-risk patients among those with diabetes.